Red Hat Gluster Storage

Red Hat Gluster Storage is the dug-out software-defined platform of storage for work with unstructured data in physical, virtual and cloud environments.

A system supports data access in the form of files and objects, has an open scalable architecture and provides economic and effective support of terabyte data arrays^[1].

It is transparent integrating physical, virtual and cloud resources, Red Hat Gluster Storage helps to transform the Big Data half structured and unstructured data to a data asset of the organizations.

The chart - representation of interaction with data storage resources by means of Red Hat Gluster Storage (2015)

Red Hat Gluster Storage is applied to storage of unstructured data of type:

• files of multimedia (images, video, audio);
• images of backup copies and operational archives;
• Big Data (the log files given to RFID and other computer-generated data);
• images of virtual machines.

The Red Hat Gluster Storage platform is based on Red Hat Enterprise Linux OS and provides deployment of cost-efficient storage systems with high availability, scalability and performance. The solution supports a number of data access protocols at the level of files and objects and also is designed counting on servers of general purpose with architecture of x86.

Preparation for start of Red Hat Gluster Storage can be executed a corporate data processing center, in public and hybrid cloud environments. The solution is optimized for corporate loadings with intensive data processing:

• archiving and backup,
• libraries of media content,
• solutions of the class "corporate DropBox",
• cloud and business applications,
• storage subsystems in virtual and cloud infrastructures,
• combination of applications on a node (korezidention)
• processing of Big Data means of Hadoop.

Red Hat Gluster Storage eliminates isolation barriers between separate storage systems, unifying processes of providing and management of local, virtual and cloud resources of storage.

Features

• Uniform global namespace
• Consolidation of memory and resources of storage in the uniform entrusted pool of resources
• Object access to a file store API interfaces for work with contents of storage as with objects
• Synchronous replication within DPC and asynchronous replication for disaster recovery
• Pictures of the file system work at the level of a cluster and allow end users to recover files easily
• Algorithm of elastic hashing
• There is no server of metadata – there is no uniform point of failure and bottlenecks
• Convenient management
• Powerful and clear tools of the command line for administrators of Linux
• Nagios
• Increase and reduction of the size of storage on the fly
• Support of standards of access
  • NFS and SMB for access to files
  • OpenStack Swift for access to objects
  • the Client of GlusterFS for high-performance parallel access

• Integration with Red Hat Enterprise Virtualization
• Centralization of resources of storage and the unified management of environments of storage and virtualization using the Red Hat Enterprise Virtualization Manager console
• Live migration of virtual machines
• Deep integration with Hadoop
  • the HDFS compatible file system eliminates costs for data movement
  • there Is no uniform point of failure
  • Data loading using NFS and FUSE

System Requirements

• Intel processor x86-64 of Xeon
• Not less than 16 GB of the RAM
• Not less than 50 GB on a disk for the system software
• Network adapter 1GbE or 10GbE
• Up to 60 disks of SAS or SATA for data storage on one RHS node
• The RAID controller with the battery of back supply and support of RAID-6 and RAID-1+0 modes.
• Contains all necessary for software for deployment on "naked iron" or virtual servers (Red Hat Enterprise Virtualization or VMware vSphere/ESXi)

2016: Gluster Storage will be unrolled on Google Cloud Platform

On February 19, 2016 the Red Hat company announced expansion of partnership with Google and the possibility of deployment of storage system of Red Hat Gluster Storage on the Google Cloud Platform platform became its result. The purpose - the organization of scalable access to files^[2] in the hybrid and public^[2].

Representation of embedding of Red Hat Gluster Storage in analytics of Hadoop (2015)

Unification of local and cloud computing of storage, "painless" transfer of the application in "cloud" allow to regulate flexibly use of cloud resources according to requirements of the enterprise. One of examples: integrate several disks Google Persistent Disk in the protected pool of storage of high availability which size can increase quickly to several petabyte in a few minutes.

The solution provides fast transfer of applications and data in "cloud" or expansion of own infrastructure of storage of the enterprise with connection of cloud services. Application of POSIX - a compatible distributed file system of Red Hat Gluster as the universal platform of storage allows to transfer various applications to the Google Cloud Platform platform without completion of a program code.

Using replication of data between DPC and Google Cloud Platform public cloud services can work as a backup system and disaster recovery for corporate applications and data.

Storage systems are beyond the traditional hardware focused approaches as the enterprises create more and more actively and implement applications based on private, hybrid and public cloud environments. Modern applications need flexible storage systems which are easily scaled together with requests of business and are capable to cope with petabytes of data in different environments. We are glad to announce expansion of partnership with Google company therefore customers of Red Hat Gluster Storage can use this storage system both in the DPCs, and in hybrid cloud environments on the basis of Google Cloud Platform. Rank Rangachari (Ranga Rangachari), vice president and general storage systems and solutions Big Data, Red Hat manager

Notes